~aleteoryx/gloss

ed42056faa25b1a8d12605ddb112e3c90fa91a28 — Aleteoryx 3 months ago
blam
5 files changed, 301 insertions(+), 0 deletions(-)

A gloss.py
A testout/a.html
A testout/b.html
A testsrc/a.gls
A testsrc/b.gls
A  => gloss.py +240 -0
@@ 1,240 @@
#!/bin/env python3

import html
import re
from datetime import datetime
from os import stat
from glob import glob
from typing import List, Optional, Union, Dict, Set, Tuple
from dataclasses import dataclass
from sys import argv, stderr, exit

usage = f'''
usage: {argv[0]} <SRCDIR> <OUTDIR>

For each .gls file in SRCDIR, creates a matching .html file in OUTDIR.

If SRCDIR/template.html exists, it will be used. Templates are
interpolated as in str.format, and the following keys are provided:

	{{title}} - The first name for an article.
	{{slug}} - The name of the output file, minus '.html'.
	{{body}} - The HTML content of an article.
	{{modtime}} - The datetime of the article's last edit.
'''[1:]

def die(why, code=1):
	print(why, file=stderr)
	exit(code)


### TYPES ###

@dataclass
class Block:
	ty: str
	text: str
	meta: str

@dataclass
class GlsFile:
	slug: str
	names: Set[str]
	blocks: List[Block]
	see_also: Optional[List[str]]


### PARSING ###

def readlines(fp):
	return filter(
		lambda x: not x.startswith(';'), 
		map(
			lambda x: x.strip(), 
			fp.readlines() ) )

def first_pass(slug, fp):
	lines = readlines(fp)

	names = set()
	for name in lines:
		if len(name) == 0 and len(names) != 0:			# section end
			break
		names.add(name.strip())
	
	if len(names) == 0:
		print(f'Empty file: {slug}.gls', file=stderr)
		return None
	
	do_see_also = False
	block_type = None
	block_text = None
	block_meta = ''
	blocks = []
	for line in lines:
		if len(line) == 0 and block_type is not None:		# block end
			blocks.append(Block(block_type, [html.escape(block_text)], block_meta))
			block_type = None
			block_text = None
			block_meta = ''
			continue
		if line.startswith('***') and block_type is None:	# see also section
			do_see_also = True
			break

		if line.startswith('~'):
			block_meta = line[1:].strip()
		elif block_type is None:
			if len(line) == 0:
				continue
			if line.startswith('>'):
				block_type = 'quote'
				block_text = line[1:].strip()
				continue
			
			if line.startswith("\\"):
				line = line[1:]
			block_type = 'para'
			block_text = line
		elif block_type == 'quote':
			if line.startswith('>'):
				line = line[1:].strip()
			block_text += "\n" + line
		elif block_type == 'para':
			block_text += ' ' + line
	
	see_also = None
	if do_see_also:
		see_also = [*filter(len, lines)]
	
	return GlsFile(slug, [*names], blocks, see_also)

quote_pat = re.compile("((?:(?!@@)(?!//).)+)(?:@@((?:(?!//).)+))?(?://(.+))?")


### GENERATION ###
	
@dataclass(init=False)
class Indexes:
	by_slug: Dict[str, GlsFile]
	by_name: Dict[str, GlsFile]
	names_sorted: List[Tuple[str, re.Pattern]]

	def __init__(self, files):
		self.by_slug = {}
		self.by_name = {}
		self.names_sorted = []
		
		for file in files:
			self.by_slug[file.slug] = file
			for name in file.names:
				if name in self.by_name:
					die(f'Redefinition of "{name}": occurs in {self.by_name[name].slug}.gls and {file.slug}.gls')
				self.by_name[name] = file
				
				pattern = re.compile(f"((?<=\\W)|^){re.escape(name)}((?=\\W)|$)", re.IGNORECASE)
				self.names_sorted.append((name, pattern))
		
		sorted(self.names_sorted, key=lambda x: len(x[0]))
			
def gen_inner_html(file, idx):
	blacklist = set()
	for name, pat in idx.names_sorted:	# populate local links
		if name in file.names:		# ...but not to the same file
			continue
		if name in blacklist:		# ...and don't repeat!
			continue
		for block in file.blocks:
			for i in range(len(block.text)):
				seg = block.text[i]
				if type(seg) != str:
					continue
				m = pat.search(seg)
				if m is None:
					continue

				s, e = m.span()
				block.text.pop(i)
				block.text.insert(i, seg[e:])
				block.text.insert(i, (f'<a href="{html.escape(idx.by_name[name].slug)}.html">{seg[s:e]}</a>',))
				block.text.insert(i, seg[:s])
				break
	
	content = f"<h1>{html.escape(file.names[0])}</h1>"
	for block in file.blocks:		# ok generate the html
		text = ''.join(map(lambda x: x if type(x) == str else x[0], block.text))
		if block.ty == 'para':
			content += f"\n<p>{text}</p>"
		elif block.ty == 'quote':
			sauce, date, url = map(
				lambda x: x if x is None else html.escape(x),
				quote_pat.match(block.meta).groups() )

			content += "\n<div>\n\t<blockquote" + ('' if url is None else f' cite="{url}"') + '>'
			content += text
			content += '</blockquote>'
			
			if sauce is not None:
				content += "\n\t<p>&mdash; <cite>" + (sauce if url is None else f'<a href="{url}">{sauce}</a>') + "</cite>"
				if date is not None:
					content += f', {date}'
				content += '</p>'
			content +='\n</div>'
	
	if file.see_also is not None and len(file.see_also) != 0:
		content += "\n<hr>\n<h3>See Also:</h3>\n<ul>"
		for slug in file.see_also:
			name = idx.by_slug[slug].names[0]
			content += f"\n\t<li><a href=\"{html.escape(slug)}.html\">{html.escape(name)}</a></li>"
		content += "\n</li>"
	
	return content


### ENTRYPOINT ###

if __name__ == '__main__':
	argv = argv[1:]
	if len(argv) != 2 or argv[0][0] == '-' or argv[1][0] == '-':
		die(usage)

	srcdir = argv[0]
	outdir = argv[1]

	try:
		with open('{srcdir}/template.html', 'rt') as fp:
			template = fp.read()
	except Exception:
		template = \
'''<!doctype html>
<html>
<head>
	<meta encoding='utf-8'>
	<title>{title}</title>
</head>
<body>
<main>
{body}
</main>
<footer>File last modified {modtime:%a, %Y-%d-%M %H:%M:%S %Z}</footer>
</body>
</html>
'''

	files = []
	for fn in glob('*.gls', root_dir=srcdir):
		with open(f'{srcdir}/{fn}', 'rt') as fp:
			data = first_pass(fn[:-4], fp)
			if data is not None:
				files.append(data)
	
	indexes = Indexes(files)
	for file in files:
		with open(f'{outdir}/{file.slug}.html', 'wt') as fp:
			ctx = {
				'title': html.escape(file.names[0]),
				'slug': html.escape(file.slug),
				'body': gen_inner_html(file, indexes),
				'modtime': datetime.fromtimestamp(stat(f'{srcdir}/{file.slug}.gls').st_mtime)
			}
			fp.write(template.format(**ctx))

A  => testout/a.html +23 -0
@@ 1,23 @@
<!doctype html>
<html>
<head>
	<meta encoding='utf-8'>
	<title>The Letter A</title>
</head>
<body>
<main>
<h1>The Letter A</h1>
<p>this is a paragraph block. these lines will be folded into one string and ultimately rendered roughly the same in the browser.</p>
<div>
	<blockquote>this is a quote block</blockquote>
	<p>&mdash; <cite>aleteoryx</cite></p>
</div>
<hr>
<h3>See Also:</h3>
<ul>
	<li><a href="b.html">The Letter B</a></li>
</li>
</main>
<footer>File last modified Thu, 2025-14-40 18:40:06 </footer>
</body>
</html>

A  => testout/b.html +19 -0
@@ 1,19 @@
<!doctype html>
<html>
<head>
	<meta encoding='utf-8'>
	<title>The Letter B</title>
</head>
<body>
<main>
<h1>The Letter B</h1>
<p>preceeded by <a href="a.html">the letter a</a>.</p>
<hr>
<h3>See Also:</h3>
<ul>
	<li><a href="a.html">The Letter A</a></li>
</li>
</main>
<footer>File last modified Thu, 2025-14-40 18:40:52 </footer>
</body>
</html>

A  => testsrc/a.gls +13 -0
@@ 1,13 @@
The Letter A

this is a
paragraph block. these lines will be folded into one
string and ultimately rendered roughly
; differently
the same in the browser.

> this is a quote block
~aleteoryx

***
b

A  => testsrc/b.gls +6 -0
@@ 1,6 @@
The Letter B

preceeded by the letter a.

***
a